Skip navigation

Birth defects after maternal exposure to corticosteroids

Corticosteroids are first-line drugs for treating a variety of conditions in women of childbearing age, but information concerning how these drugs affect pregnancy is limited. This paper conducts a study and a meta-analysis to determine the risk of major malformations, and more specifically oral clefts, of corticosteroid use on the foetus.


The risk of oral clefts increases threefold in foetuses exposed to corticosteroids during the first trimester.


L Park-Wyllie et al. Birth defects after maternal exposure to corticosteroids: prospective cohort study and meta-analysis of epidemiological studies. Teratology 2000 62: 385-392.


Participants were women who had voluntarily contacted a counsellor, from the Canadian Motherisk Program, during pregnancy. One group of 184 women had been exposed to prednisone (a synthetic corticosteroid) during pregnancy. The unexposed group of 188 women had enquired about either topical retinoic acid for uncomplicated acne or oral astemizole for seasonal allergies (neither of which has been associated with an increased risk of major malformations).

Information was collected on the mothers' medication, obstetric, medical and genetic history and drug exposure. One year after the expected date of delivery details were collected on the outcome of the pregnancy, birth weight, presence/absence of birth defects and perinatal and neonatal complications (which were corroborated by the child's physician).

In the exposed women, prednisone was administered for an average of 21 weeks (SD 16); the average daily dose was 27 mg (SD 29); and 75% were exposed to it in the first trimester of pregnancy.

The number of major birth defects after corticosteroid use was recorded. Major defects were defined as life-threatening, requiring surgical intervention or having serious cosmetic ramifications. Cases were excluded if factors other than the corticosteroids could have caused the malformation (i.e. genetic syndromes or maternal infections).

Study Results

Compared with the unexposed women, women exposed to corticosteroids were more likely to be smokers. Alcohol consumption and age (average 30 years) did not differ between the two groups.

The number of live-born babies was similar in the exposed and unexposed groups (157 and 171 respectively). There were no differences in the number of miscarriages between the two groups (13 in each), foetal deaths (one in each) or stillbirths (one in unexposed group). The number of elective terminations was higher in the exposed group (16 versus two). Babies born to exposed mothers were smaller (mean 3,112 g versus 3,428 g) were born earlier (mean 38 weeks versus 39.5 weeks) and more likely to be premature (27 babies versus nine). Despite these differences, the majority of babies in both groups were an appropriate weight for their gestational age.

There was no difference in the number of major abnormalities between the groups (four out of 111 exposed babies; three out of 172 unexposed babies). (Only foetuses exposed to corticosteroids during the first trimester of pregnancy were included in this analysis.)


The databases MEDLINE (from 1966 to 1999), Embase (from 1988 to 1999) and Current Contents (1999) were searched. Bibliographies from retrieved papers were then reviewed. Controlled studies examining first-trimester systemic exposure for any corticosteroids, any dose, all conditions, any duration and all languages were included. Studies with fewer than ten patients and those which examined topical or inhaled steroids were excluded.

Search Results

Ten studies met the inclusion criteria (six cohort conducted from 1962 to 2000) and four case-control conducted from 1994 to 1999). The studies included women with various diseases including rheumatoid arthritis, asthma and inflammatory bowel disease. Participant numbers ranged from 22 to 50,282 in the cohort studies (totalling 51,470) and 1,396 to 56,557 in the case-control studies (totalling 71,705). Four examined the effects of corticosteroids and other medications and six examined only corticosteroids.

In the analysis of the six cohort studies, corticosteroid exposure was not associated with an increased risk of major malformations. Because the largest study (50,282 participants) did not distinguish between major and minor malformations and the purpose of this analysis was to examine major defects, the analysis was repeated without this study. A significant risk of major defects with corticosteroid exposure was then found (see Table 1). In the studies that specified the malformations, cleft palate was the most common defect with three cases among 390 corticosteroid exposed foetuses compared with no cases among 708 non-exposed foetuses.

Table 1. Risk of major malformations from cohort study analysis with and without largest trial.

Cohort Studies



Odds ratio (95% confidence interval)

Six studies

16 cases out of 535

2,275 cases out 50,845

1.45 (0.81 to 2.60)

Five studies

10 cases out of 390

4 cases out of 708

3.03 (1.08 to 8.54)

The four case-control studies examined the risk of oral clefts. There was an increased risk of oral clefts when the foetus was exposed to corticosteroids during the first trimester (odds ratio 3.35, 95% confidence interval 1.97 to 5.69).


Although this paper focuses on major malformations, the differences in pregnancy/baby characteristics between exposed and unexposed mothers in the study are interesting. The reason why 16 exposed women chose to terminate their pregnancies after prednisone exposure may be due to the suggestion that prednisone during pregnancy may cause oral clefts, but this is a speculative suggestion.

There was not an increased risk of major malformations either in the study or the meta-analysis without the largest cohort trial (leaving mostly small and/or old cohort studies), but the meta-analysis of case-control studies showed a greater than threefold increase in the risk of oral clefts in foetuses exposed to corticosteroids. Labelling a medication as the cause of a major birth defect has serious clinical implications. The evidence that corticosteroids increases the risk of oral clefts has to be balanced against potentially serious implications for the mother and indirectly to the foetus if steroid treatment is discontinued or not initiated, but it allows clinicians to make an informed decision concerning the use of corticosteroids in pregnancy.